
	Weekly SDAST chat #55: Thursday 10th January 2002
	
Participants: Peter Kretschmar, Carol Anne Oxborrow, Niels Joergen Westergaard,
	Stefan Larsson and Sami Maisala
	
Subjects discussed: 
	-ARF data structures and templates
	Status/schedule for SVT-E and PLC
	Problems with j_calib tools
	Redelivery of j_src_spectra
	Status/schedule for Science Verification
	Installing scripts and ISSW
	Using RIL/PIL interactively
	AI List
	Getting Extra capacity on ISDC accounts
	
Action Items: no new AIs

**** Logging Started : Thu Jan 10 09:56:57 MET 2002
> Hello Chaps - everyone survived Christmas? END
<larsson> Very well! END
<njw> Morning END
> We'll start as soon as Peter arrives: I'm assuming that Sami hasn't got
> any of me emails about the chat. END
*** peter (~pkretsch@crab.unige.ch) has joined channel #jemxadr
<peter> Good Morning everyone! END
> Hi Peter - happy New Year to you, and all the SDAST chaps! END
<njw> Morning, Peter END
<larsson> Hi Peter. END
<peter> Sami told me, that 10:00 would not be a problem, let's see. END
> So we'll get started. Just a quick note to Peter: I'm working on the
<peter> To CAO: Thanks and happy new year to you too! END
> j_calib_gain_fitting problem, but unfortunately I've been off sick for
> two days, so no progress on it whie you and Tess were busy exchanging
> emails. END
<peter> Yes, thought so - otherwise you would have surely responded. Hope you 
feel better now. END
> We'll wait for Sami, if he says he's coming. Thanks Peter, I'm down to
> blowing my nose only every half hour now. END
<peter> Maybe we can sort out something between Niels Jrgen, Stefan and me 
first:
<peter> In my opinion we should NOT use the name JMXi-SRCL-ARF for the formal 
on-axis ARF, because that name implies ARF's specific to sources.
<peter> Ehm, actually we have two such templates alread -SLSP-ARF (source list 
spectra) and -SLLC-ARF (source list LC).
<peter> I _think we have the two because they are generated by 
j_src_[spectra|lc] independently - is that correct Stefan?
<peter> END
<larsson> Since we will always have the same number of energy bins in the ARF 
and 
<larsson> will do rebinning afterwards as needed
<larsson> that is acually not necessary anymore. END
<njw> OK, Peter, we can very well set up a new name for the standard (on-axis)
<njw> I'm awaiting Marc's reaction END
<njw> Sorry, 'ARF' fell out from the first line of my comment END
<larsson> (To Peters question: Yes) END
<peter> Just as a note: ISDC's Configuration Control Board (CCB - an acronym to 
know and fear :-) ) plans to put the V2 templates
<njw> Stefan, are you suggesting to merge -SLLC-ARF and -SLSP-ARF to a single 
name ? END
<peter> under configuration control _this_ Friday. So if we do the merging, we 
need to move fast. END
<larsson> Maybe that will create problems. I would prefer one
<larsson> but if we should avoid such changes I can live with the two. END
<peter> Let's keep it as it is for the moment and make the change later. OK? END
<larsson> OK. END
<peter> Btw, I have Sami on the phone - he has connection problems. END
> Lets give him 5 minutes to get sorted out, then start on the agenda END
*** smaisala (~smaisala@cetus.astro.helsinki.fi) has joined channel #jemxadr
<peter> Hi Sami!
> Hi Sami - Happy new Year! END
<smaisala> Hi all! After a little technical problems I'm here. END
<njw> Hi Sami END
<smaisala> Happy new Year too! END
> Let's start on point one: status for SVT-E and PLC: how's it going Peter,
> and what've we GOT to be finished with and when? END
<peter> On ISDC's side it _appears_ mostly OK (we'll know more after tomorrow's 
dry run)
<peter> Since we have some problems with the j_calib_ tools (normal for 
something tested for the first time), as you know from Tess' and my emails, they 
will NOT be part of the S/W executed  in SVT-E
<peter> as far as I understand - only well-tested software is included.
> I'll try to get the j_calib problems cleared up today and tomorrow. END
<peter> Well, Tess is supposed to deliver tomorrow, so maybe if all goes 
quickly. But we also need the formal IC delivery of an IMOD-GRP with the 
ECAL-MOD for j_calib_adc.
> Yes, if it's got to be done properly, those things will only be ready for
> PLC. END
<peter> In any case, if we cannot run them within SVT-E itself, ISDC will 
exercise them during PLC and later in SVT-E reruns etc.
<peter> So no reason to slow down on these, but it may be too late for the 
actual SVT-E. END
> I understand that Peter. What about other fronts of JEMX ISSW, all okay
> there? END
<peter> Not really - I'm still waiting for one more redelivery of j_src_spectra, 
which is waiting for the formal response file, which is currently being sorted 
out by a lot of activity.
<peter> Niels Jrgen has been very productive these days but it seems we always 
find one more issue that needs closing first.
> So that's all that's outstanding for SVT-E/PLC: j_calib tools, 
<peter> There is a lot of pressure on me here to have the scripts running. 45 
minutes ago, I promised to have something next week and VERY much hope that this 
really works! END
> j_src_spectra and the response file? END
> While only you can judge how many loose ends there are for the scripts,
<peter> In _principle, j_src_spectra and the response file are not required for 
SVT-E/PLC but OTOH we would naturally love to test the analysis software also 
with this. END
> it seems like things are more or less under control for the rest of ISSW.
> Any comments? END
<njw> N
<smaisala> n
<larsson> n
<peter> The scripts themselves are in quite good shape. I'd do only minor 
changes before a delivery. My main problem there would be the effort to describe 
EXACTLY to Tess what has chnged sincethe last formal delivery. END
> Okay, so next item: where are we with the Science Verification? END
<peter> Well, NJW sent me his inputs and now it's my problem to:
<peter> - find time to set up the OSim input
<peter> - make sure we have a useful OSim for JEM-X purposes
<njw> I have made a 'testsky' with sources arranged so that the cases can be 
tested
<njw> by appropriate pointings. END
<peter> - run (or have Ada run) the simulator to generate test data sets.
<peter> Btw, this is very nice work by you Niels Jrgen and the Science Analysis 
Meeting was quite impressed.
<peter> There will be probably pressure  on IBIS to do something similar. END
> Good stuff NJW! Keep up the good work! END
<peter> (OTOH onece I have set up the test catalogue, we could reuse it for IBIS 
in some sense). END
<njw> Mostly to Peter and Stefan: I think I have a new version of j_os_raytrace
<njw> that includes the new event handling compatible with the latest version of
<njw> the RMF. I-m doing some control checks today. END
<larsson> Very good NJ. END
> I'll wait till final delivery of the scripts for SVT-E/PLC before installing
> them here: do I need OPUS or something for this? END
<peter> No, the scripts can be installed like any executable (configure, make 
global_install) and then be run manually.
> Well that's good news. And I also have to install ALL the ISSW? ARGGHH!END
<peter> The OPUS system is used in this case only to decide the CPU to run them 
on and to automatically launch things like Science Window QLA where many, many  
tasks are started in short time intervals. END
> I'll look forward to all of that somewhat later in the month.END
> Any more comments on Science verification? END
<peter> Naturally :-) But it may be actually easier than you think. It could 
well be that running the typical commands in a parent directory of all the ISSW
<peter> will nicely loop over all the things. I'll try and ask Mathias what is 
required to have this nice package installation working everywhere (it sure 
works here). END
> Thanks Peter. Another question you should put to Mathias: what about
<njw> One comment on the science verification: If we can agree on the testsky 
and the
> horrible/cheap/quick user interfaces with printf/scanf etc. for V3 tools?END
<njw> pointings and the exposure times (not defined yet) we can do the 
simulations
<njw> in Stockholm, here and at ISDC. Then we can compare results and enhance 
the
<njw> statistical weight of the results. END
<peter> OK, let me first answer Carol Anne's question:
<peter> We've talked briefly and first he said "why not" but then after a bit of 
reflection noted that one probably can use PIL and RIL also in an interactive 
way
<peter> - there is nothing keeping you from, e.g., asking for certain parameters 
only if others have been set or for going back to the same PIL call over and 
over.
> That sounds like a horrible hack......END
<peter> And if this _is_ sufficient (too be confirmed) then using these 
libraries could make it easier to interact with standard GUI's etc later.
<peter> In brief: he'd prefer RIL and PIL, if this is possible (TBC), but if you 
have good arguments, basic C functions can be used for interactive software that 
is not called by scripts or pipelines.
<peter> One more argument for PIL/RIL - your tool could then potentially called 
from some person's script in a few years. END
<peter> Maybe we two can discuss what you want to do and see if RIL/PIL could 
make sense or not. END
<peter> Now to Niels Jrgens point:
> Okay, and NJW's question? END
<peter> A very good idea, BUT we here first need to get the OSim updated so we 
all can generate valid test data directly. 
<peter> But once this is done (maybe a bit of work still), then this should be 
very useful. END
<peter> One thing to note, though: we will need to come to a coherent set of 
conclusions that will translate to tools being accepted, or (more probably) SPRs 
and SCREWs to be written. END
<njw> Alright, but Stefan and I can start to get a first estimate of e.g. 
exposure times END
<larsson> I agree. END
> Any more comments on Science Verification? END
<njw> N
<peter> Not from me. 
> Okay, on to the AI list: As far as I can see a few more have been closed
> namely 21.1 and 21.3: other than that I don't see much change. Any comments?
> Any closed AIs I don't know about? END
<peter> Nothing here (I think I have no open action, besides "All") END
<njw> Just a question Carol Anne: is the ~oxborrow/sdast/AIlist.txt the updated
> What about 21.6 NJW that's been discussed above, when will that be closed? END
<njw> version of the list? END
> No, if you get to the list via the chat page you end up at the very pretty
> html version (colour designed to be action-inspiring!)
> ~oxborrow/sdast/AIlist.html
> END
<njw> I see, yes, the RMF has been given to Stefan so in fact it is closed END
> Good stuff, I'll move it over to the even prettier AIlist.closed.html
> (colour designed to be smugness-inducing!) END
> So is that all for the AI list? END
<larsson> Y fr. me END
<njw> Y
<peter> Y
<smaisala> y
> HO HO HO! ANY OTHER BUSINESS? END
<njw> N
<larsson> No business like show business. END
<smaisala> To Peter: Is there any possibilities to get more disc space from ISDC
? (quota 200000) END
<peter> To Sami: In principle yes. But what you probably should do is move stuff
 to our "unsaved" disks.
<peter> It's probably best if you write a nice email to Alain Aubord and tell 
him your problem. END
<smaisala> Won't help, I can't extract test data to my executable. END
<smaisala> OK: END.
> Any more for any more? END
<smaisala> N
<larsson> No. END
<njw> N
<njw> Bye, have a good day END
> In that case we can pat ourselves on the back (or have your office mate do it)
*** njw has left channel #jemxadr : (njw)
> for finishing in an hour: good bye and have a good weekend. END
<smaisala> bye 
*** Signoff: smaisala (Client Exiting)
